A Deepfake Simulation is set up like a phishing simulation, in a short wizard. This article walks you through the setup in the SoSafe Manager. For an overview of how Deepfake Simulation works, see Introduction to Deepfake Simulation.
Before you start
- You need a deepfake for the scenario you want to use. A colleague has to give their consent and record their photo and voice before you can create the simulation. This can take some time, so start early. See Request consent in the Character Library.
- Set up whitelisting. As with phishing simulations, the simulation only starts after you've confirmed that whitelisting is done. The domains for the meeting page and the chat link are listed under Settings / Whitelisting, together with the other simulation domains. See Whitelisting.
- You need the right admin role. Administrators and Campaign Managers can create Deepfake Simulations. See Admin user roles and permissions.
Create the simulation
- In the SoSafe Manager, go to Simulations / Simulation Manager.
- Open the Deepfake tab and select Create deepfake simulation.
If you haven't created a Deepfake Simulation yet, the Deepfake tab shows the available scenarios with an example video and summarizes what happens for you and for your employees. Once you've created simulations, it lists them instead.
You can also start from the Character Library: on the details page of a ready deepfake, select Create deepfake simulation.
The setup is a 4-step wizard: Scenarios, Audience, Timeframe and Reporting method, followed by a summary. Use Continue and Back to move between steps. Leave exits the wizard.
1. Scenarios
Choose the scenario you want to send. Each simulation can only include 1 scenario. Each scenario shows its difficulty, the colleague whose deepfake is used in it and the language the deepfake speaks.
To see a scenario in detail, select the eye icon. The Deepfake scenario preview shows the invitation email, the deepfake's language, the difficulty, the interaction and the attack types. Select the link in the email to join the simulated call and test the whole flow, just as your employees will experience it.
2. Audience
At the top, you see the language the selected scenario is delivered in. Then choose who receives the simulation:
- All users sends the simulation to all your users. New employees added to your user data are included automatically. The badge shows how many of your users have the scenario's language as their language, for example 4/12 German.
- Target specific groups lets you select teams, departments or high-risk groups.
You can also send the simulation to users with a different language. Each deepfake speaks only 1 language, so these users hear it in the scenario's language.
The CEO Fraud scenario is intended for employees who can approve or release payments, for example in Finance. We recommend selecting Target specific groups and sending it to these groups only.
3. Timeframe
Set the Start date and End date. The invitation emails are sent at different times within this timeframe. We recommend keeping the simulation to 4 weeks.
Make sure the consent for the deepfake is valid for the whole timeframe. If the consent expires, the simulation is deactivated and archived.
4. Reporting method
Decide how much detail you want in Analytics:
- Individual shows which employees interacted with the simulation.
- User group based (recommended) shows results per user group. You can't identify individual employees.
You can't change this setting once the simulation has started. You can choose a different reporting method for future simulations.
5. Summary
The Deepfake summary shows all your settings. Give the simulation a name (it defaults to "Simulation" plus today's date) and check the completed steps. To change one, select the pencil icon next to it. When everything looks right, select Done.
After you create the simulation
You're taken back to the Deepfake tab. The new simulation is listed under Pending with the label Whitelisting. Update your whitelisting settings under Settings / Whitelisting, then select Whitelisting done to launch the simulation.
The simulation is then listed under Live, or under Scheduled if its start date is in the future. To view, change or stop it, open the menu with the three vertical dots on its card and select See details, Edit or Deactivate.
To follow the results, see Deepfake Simulation metrics.
What your employees see
- Employees receive an invitation email that asks them to join a short call. It contains a link that looks like a Microsoft Teams meeting link.
- If they select the link, a page that looks like Microsoft Teams opens in the browser, with a meeting with your colleague. To join the call, they select Join now.
- In the call, the deepfake video plays. The deepfake makes an urgent request and posts a link in the Meeting chat.
- If they select the link in the chat, they land on a learning page. It explains that the call was a simulated deepfake attack, created with the colleague's consent, and shows the warning signs in the email, the lobby and the call.
Everyone who joins the call receives an email after the video ends. It explains that the call was a simulation and that the person in the video was an AI-generated deepfake of a colleague who agreed to it. It also lists how to spot a real deepfake attack. Employees who didn't join the call don't receive this email.